A Fuzzy Multi Agent Architecture for Sensor based Automated Scene Surveillance
Najar Yosra, Raouf Ketata, Zrill Rym, Hamrouni Mahmoud, Ksouri Mekki · J. Multim. Process. Technol. · 2013
Human recognition and movement analysis are important tasks in various applications especially in video surveillance. This paper proposes a multi agent architecture for sensor based automated scene surveillance using Microsoft Kinecet camera. It aims to record videos about each person that appears in the scene and to generate sonorous alarms when an abnormal situation is deduced. This decision making support system had to deal with environment uncertainties through fuzzy expert system which is responsible of person recognition.